For all the benefits virtualization brings to a data center – reduced rack footprint and power consumption, greater efficiency and reliability, lower overall costs – enterprises frequently leave other benefits of virtualization on the table by not adapting the data center physical infrastructure to the dynamic new IT environment.
In a related blog post, we discussed several elements crucial to realizing the full potential of virtualization in a data center.
One is row-based cooling. Because virtualization results in dynamic and high-density IT loads, with changing “hot spots,” row-based cooling has proven to be an effective method in these types of data centers. This method of cooling relies on proximity and instrumentation to sense and respond to server temperature changes.
Another element is scalable power and cooling. While virtualization reduces power consumption, it increases power inefficiency in data centers that have reduced IT loads because power and cooling devices are less efficient at lower loads.
Having a scalable power and cooling system allows data center owners to reduce fixed costs and increase efficiency.
Eyes and ears of the data center
The third element, and the focus of this post, is capacity management. With dynamic IT loads and scalable physical infrastructure, virtualized data centers are all about constant change. Given that so much is at stake in the event of downtime, it is paramount for such a dynamic environment to be monitored in real time.
Basically, capacity management serves as the eyes and ears of your virtualized data center team. Capacity management provides the software and instrumentation for real-time monitoring and analysis of data center power, cooling and physical space. The idea is to ensure you have adequate amounts of each, even as IT loads vary over time and location.
Capacity management ensures this because it is able to see, in real time and at the rack and server level, where physical infrastructure resources are needed for various IT loads, where capacity is dangerously low, and where there is unusable (known as “stranded”) capacity.
It is capacity management that allows a data center to manage change, whether it’s server population, power density, load migration or the implementation of new technologies. Conversely, unmanaged change can lead to downtime and wasted resources.
Beyond real-time system monitoring, capacity management also can help with planning and strategy, using automated intelligence and modeling to suggest the optimal place for locating equipment, predict the impact of proposed changes, and flag trends headed toward bad outcomes.
A capacity management system that computes server locations and loads, power and cooling capacity available to servers, variations in room temperatures, and power consumption is essential to a stable and reliable virtualized data center. It can detect unseen changes in location and demand for power and cooling resulting from virtualization, and can identify complex interdependencies between power, cooling and space capacities.
Bottom line: An effective capacity management system optimizes the use of available resources, thus improving data center efficiency.
To learn more about maximizing the benefits of virtualization, check out the APC by Schneider Electric white paper, Virtualization: Optimized Power and Cooling to Maximize Benefits.